首都医科大学学报 ›› 2018, Vol. 39 ›› Issue (1): 128-132.doi: 10.3969/j.issn.1006-7795.2018.01.022

• 临床研究 • 上一篇    下一篇

应用空腹1 h尿钙浓度能否诊断糖尿病患者高尿钙?

刘峥1, 张鹏睿2, 陈哲2, 高霞2   

  1. 1. 首都医科大学石景山教学医院北京市石景山医院门诊办公室, 北京 100043;
    2. 首都医科大学附属北京朝阳医院内分泌科, 北京 100020
  • 收稿日期:2017-02-22 出版日期:2018-01-21 发布日期:2018-01-27
  • 通讯作者: 刘峥 E-mail:948169558@qq.com

Can urinary calcium concentration after 1 hour fasting be used to diagnose hypercalciuria in diabetes mellitus patients?

Liu Zheng1, Zhang Pengrui2, Chen Zhe2, Gao Xia2   

  1. 1. Outpatient Office, Shijingshan Teaching Hospital, Capital Medical University, Beijing Shijingshan Hospital, Beijing 100043, China;
    2. Department of Endocrine, Beijing Chaoyang Hospital, Capital Medical University, Beijing 100020, China
  • Received:2017-02-22 Online:2018-01-21 Published:2018-01-27

摘要: 目的 探索能否在糖尿病患者中应用空腹1 h尿钙浓度诊断高尿钙,寻找24 h尿钙(24 hour urine calcium,24 hUCa)的影响因素,建立回归方程,并判断清晨空腹1 h尿钙浓度(1 hour urine calcium,1 hUCa)与24 hUCa的一致性,寻找清晨空腹1 h尿钙浓度最佳诊断切点。方法 留取61名在北京朝阳医院内分泌科住院的糖尿病患者24 h尿、随后的空腹1 h尿标本(即6时至7时尿),分别检测24 h尿钙、清晨空腹1 h尿钙浓度,计量24 h尿量,检测25-羟基-维生素D和甲状旁腺激素,应用多元回归分析,寻找24 hUCa的影响因素,建立回归方程,应用Bland-Altman分析法,将清晨空腹1 h尿钙浓度与24 hUCa进行一致性分析。将24 hUCa>350 mg/d选取为诊断切点,以清晨空腹1 h尿钙浓度为检验变量,做受试者工作特征(receiver operating characteristic,ROC)曲线,并且计算其灵敏度、特异度、误诊率。进行配对资料的χ2检验。结果 清晨空腹1 h尿钙浓度是24 h尿钙的影响因素,可建立回归方程。Bland-Altman一致性分析显示,清晨空腹1 h尿钙浓度与24 hUCa的一致性好。将24 hUCa>350 mg/d选取为诊断切点,以清晨空腹1 h尿钙浓度为检验变量,做ROC曲线时,曲线下面积为0.925,清晨空腹1 h尿钙浓度选取171.4 mg/L(即4.285 mmol/L)为诊断高尿钙的切点时,灵敏度为100%,特异度为84.9%,误诊率为50%。应用清晨空腹1 h尿钙诊断时,8例阳性结果被正常检出,但8例阴性结果成为假阳性。Kappa系数为0.596,提示一致性一般。应用McNemar方法,提示差异有统计学意义。结论 在糖尿病患者,由于误诊率高,空腹1 h尿钙浓度不能诊断高尿钙。但是,由于灵敏度高,其可作为初筛试验,阳性者再行24 hUCa检测。

关键词: 高尿钙, 诊断, 空腹1 h尿钙浓度

Abstract: Objective To explore if we can replace 24 hour urine calcium (24 hUCa) with fasting 1 hour urinary calcium (1 hUCa)concentration in diabetes mellitus patients. We explored the factors affecting 24 hUCa,and built the regression equation. Further we judged the agreement of fasting 1 hUCa concentration and 24 hUCa. Finally,we tried to find the best cut-off for fasting 1-hour urinary calcium concentration.Methods Sixty-one diabetes mellitus patients in the endocrine ward of Beijing Chaoyang Hospital, Capital Medical University were invited to attend this test.We collectd the 24 hour urine samples and the following 1 hour urine samples (fasting condition in 6 to 7 o'clock), we detected 24 hUCa, 1 hUCa, measured 24 hour urine output,detected 25-OH-vitamin D and parathyroid hormone (PTH). We did multiple regression analysis to find the factors affecting 24 hUCa,and built the regression equation. More,we judged agreement of fasting 1 hour urinary calcium concentration and 24 hUCa with Bland-Altman analysis. Finally, we made 24 hUCa >350 mg/d the cut point, and used 1 hUCa to draw receiver operating characteristic (ROC) curve, evaluated the sensitivity, the specificity, the misdiagnosis rate.Then we made the chi-square test of paired data.Results Fasting 1 hUCa concentration was the factor affecting 24 hUCa. We could build the regression equation.The agreement of fasting 1 hUCa and 24 hUCa was good. When we diagnosed hypercalciuria with 1 hUCa, choosing 24 hUCa>350 mg/d as the cut point to draw ROC curve,the area under the curve is 0.925. When 1 hUCa was 171.4 mg/L (4.285 mmol/L), it had the sensitivity of 100%、the specificity of 84.9% and the misdiagnosis rate of 50%. When we diagnosed hypercalciuria with 1 hUCa,8 true positives were found,but 8 false positives were diagnosed too. Kappa's coefficient was 0.596, that meant the agreement was ordinary.And we used McNemar method,the difference was statistically significant.Conclusion In diabetes mellitus patients,because of the high misdiagnosis rate, fasting 1 hour urinary calcium concentration can not be used to diagnose hypercalciuria. But,the sensitivity is good,thus it can be used as the primary screening test.When the result of 1 hUCa detection is positive,we should measure 24 hUCa again.

Key words: hypercalciuria, diagnosis, fasting 1 hour urinary calcium concentration

中图分类号: